Future Directions of Spine Surgical Research - Use of Machine Learning Models to Optimize Treatment and Outcome Prediction in Degenerative Spondylolisthesis

• Applicant: Dr. Annika Bay
• Subject Area: Orthopaedics, Traumatology, Reconstructive Surgery
• Term: since 2023
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 526240791

The overall objective of this research project is to investigate whether predictive model development for benchmarking and outcome prediction using machine learning (ML) algorithms can support spine surgical decision making to improve patient care. A large, multi-surgeon, prospectively maintained database of the Hospital of Special Surgery, NY, USA (n=1.083) will be used to develop machine learning models which predict outcome in patients with symptomatic single segment L4/L5 spondylolisthesis following either decompression alone or decompression with fusion surgery. Further, we aim to develop a novel classification system for spondylolisthesis built on clinical data and ML prediction to allow individualized treatment recommendations. Once all the models have been retrained, the one with the highest performance on the test dataset will be exported for hosting on a web app to be used with external patient data. We propose that by utilizing ML algorithms, novel patient-specific key variables will be identified, enabling spine surgeons to predict the best treatment option and predict outcome for patients with DS. Therefore, informed and shared decision making can be reformed using ML as a novel support tool in clinical practice.

• DFG Programme: WBP Fellowship
• International Connection: USA
• Host: Professor Dr. Sheeraz A. Qureshi